Record of the Queen Victoria Museum and Art Gallery 120 Launceston, Tasmania Record of the Queen Victoria Museum and Art Gallery No. 120 (a peer-reviewed journal) is published by the Queen Victoria Museum and Art Gallery Launceston, Tasmania, Australia 7250

Publications Coordinator: Andrew Parsons, QVMAG Photographs: Rolan Eberhard, unless otherwise labelled Design and layout: Renée Singline, QVMAG Printing: Foot and Playsted Pty. Ltd., Launceston, Tasmania 7250

ISSN 0085-5278

Cover photograph: Emu bones from Caveside (QVM.1974.2.5-7) Tasmanian emu (Dromaius novaehollandiae diemenensis) at the Queen Victoria Museum and Art Gallery, Launceston: description, provenance, age

Rolan S Eberhard

Record of the Queen Victoria Museum and Art Gallery No. 120

June 2020 Acknowledgements

This QVMAG Record is the result of a series of events which trace back to Paul Flood and Ross McNeill, who told me of some large bones they had seen in a cave at Mole Creek. The bones included those of an emu, one of the specimens discussed in this report. That discovery prompted a search of the QVMAG collection for other emus, of which we found several, including some of questionable provenance. At that stage the Plomley Foundation stepped in with funding to date a selection of specimens, underpinning the findings presented here. I thank the Plomely Foundation for its generosity. I also thank Senior Curator Natural Sciences, David Maynard, Collection Officer Natural Sciences, Tammy Gordon, and Library Coordinator, Andrew Parsons, for their advice and support throughout the project. I thank also the many other people at QVMAG and other institutions who assisted.

Abbreviations AMS accelerator mass spectrometry (radiocarbon dating) ANU Australian National University Before Present (by convention ‘present’ for dating BP purposes is 1/1/1950) calAD calibrated (years) anno Domini calBC calibrated (years) before Christ cm centimetres Department of Primary Industries, Parks, Water and DPIPWE Environment g grams ka kilo-annum (1000 years) km kilometres m metres

MIS Marine Isotope Stage

mm millimetres

QVMAG Queen Victoria Museum and Art Gallery

iii Holocene epoch, or are considered likely to Executive summary be of this age. This material is younger than In 2017 the Plomley Foundation allocated 10 000 years BP. Three samples of bone funding to review holdings of the Tasmanian were dated 895-1021 calAD, 1021-1152 calAD emu (Dromaius novaehollandiae diemenensis) in and 1043-1223 calAD. The dated bones are the collection of the Queen Victoria Museum from three partial emu skeletons collected in and Art Gallery, Launceston. The central caves around Mole Creek in central northern purpose of the project was to clarify the age Tasmania. These represent the most complete and provenance of the material, some of which skeletal assemblages of Tasmanian emu in any had been regarded as doubtfully Tasmanian. public collection. A fourth, undated specimen In total, 27 lots of registered and unregistered comprises a single weathered femur fragment material, comprising approximately 250 from Mt Cameron West on the far north west bones and bone fragments, minor pieces coast. of eggshell and a dried tissue specimen, The most recent emu specimen in the were examined. Associated documentation, collection is of broadly known age and dates including specimen labels, database entries, to the mid-19th century. This is the celebrated published and unpublished references, was dried emu leg from Ronald Campbell Gunn’s consulted. Selected specimens were dated residence at Newstead House in Launceston. using the radiocarbon method. The results of this work are presented here as a descriptive A further four registered and unregistered inventory of emu holdings at QVMAG, with lots are confirmed as emu but may or may commentary on the condition, age and not be Tasmanian. Possibly, some of these are provenance of the material. mainland emu material used for comparative purposes. An additional unregistered Some 13 registered lots comprising 17 bones specimen was found not to be emu. in total are confirmed as Late Pleistocene in age, or are considered likely to be of this Summary details for the respective lots are age. This material is older than 10 000 years provided in the table below. BP and some of it may exceed the limits of radiocarbon dating (~50 000 years BP). Three samples of bone were dated at 37 909- 33 764 calBC, 40 131-36 665 calBC and 40 799-37 138 calBC. These are the oldest dates yet obtained for emu in Tasmania. They sample the emu population at a time of colder climatic conditions and lower sea levels, when Bass Strait did not preclude mixing of emus in Tasmania and Victoria. The Pleistocene age material was collected in the Smithton area of north west Tasmania, from swamps and Scotchtown Cave. In addition to bone, the Pleistocene age material includes three small fragments of emu eggshell. These were dated 15 839- 15 421 calBC, providing hitherto disregarded evidence for the presence of emu on Flinders Island. Some eight registered and unregistered lots comprising over 200 bones and bone fragments are confirmed as dating to the

iv Summary details for the respective lots are provided in the table below.

Lot name QVMAG no. Item Locality Age Late Pleistocene QVM.2007.GFV.5 Scotchtown Cave, 37 909-33 764 Phalanges (2) Smithton calBC (SANU-56324) Late Pleistocene Scotchtown Cave, 40 799-37 138 QVM.2005.GFV.45 Tibiotarsi (2) Scotchtown Smithton calBC Cave emu (SANU-56326) Late Pleistocene Scotchtown Cave, 40 131-36 665 QVM.2006.GFV. 120 Tarsometatarsus Smithton calBC (SANU-56325) Scotchtown Cave, Undated (probably QVM.2006.GFV. 121 Femur fragment Smithton Pleistocene) Undated (probably QVM.1990.GFV.138 Tibiotarsus Mowbray Swamp Pleistocene) Undated (probably QVM.1990.GFV.139 Femur Mowbray Swamp Pleistocene) Undated (probably QVM.1990.GFV.140 Tarsometatarsus Mowbray Swamp Mowbray Pleistocene) Swamp emu Undated (probably QVM.1990.GFV.141 Tarsometatarsus Mowbray Swamp Pleistocene) Undated (probably QVM.1990.GFV.142 Vertebra Mowbray Swamp Pleistocene) Undated (probably QVM.1990.GFV.143 Synsacrum Mowbray Swamp Pleistocene) Irishtown Undated (probably QVM.1990.GFV.144 Tibiotarsus Irishtown tibiotarsus Pleistocene) Late Pleistocene Eggshell 15 839-15 421 Jackson egg QVM.1965.GFV.0006 Flinders Island fragments (3) calBC (SANU-55616) Mole Creek Undated (probably QVM.1991.GFV.54 Tibiotarsus Mole Creek tibiotarsus Pleistocene) Mole Creek Unknown Undated (probably (fractured) QVM.1489 Tibiotarsus (recorded as Mole recent)1 tibiotarsus Creek) Dromaius Cave, Holocene (see QVM.2013.GFV.11 Tibiotarsus Dromaius Mayberry below) Cave emu Partial skeleton Dromaius Cave, 1 021-1 152 calAD QVM.2016.2.008 (approx. 70 total) Mayberry (SANU-49415)

1 Here, ‘recent’ means post-European settlement of Australia. Strictly, this falls within the Holocene epoch (11,700 years BP to present).

v Lot name QVMAG no. Item Locality Age Caveside-Mole Holocene (see QVM.1974.2.5 Synsacrum Creek below) Caveside-Mole Holocene (see QVM.1974.2.6 Femur Creek below) Caveside Emu A Caveside-Mole Holocene (see QVM.1974.2.7 Ischia (2) Creek below) Skull and Holocene Caveside-Mole QVM.1974.2.9 postcranial bones 1 043-1 223 calAD Creek (approx. 66 total) (SANU-56323) Unregistered Post cranial bones Holocene Caveside-Mole Caveside Emu B (associated with (approx. 90 in 895-1 021 CALAD Creek QVM.1974.2.9) total) (sanu-52431) Mt Cameron West QVM.1993. Mt Cameron Unknown (probably Femur femur GFV.146 West Holocene) Newstead House QVM.2002.2.1 Dried lower leg ?St. Pauls Plains 19th century emu leg Unknown Anomalous ‘King (stored with Undated (probably Unregistered Femur Island’ femur King Island emu recent) bones) Unknown Anomalous ‘King (stored with Undated (probably QVM.1993.GFV.18 Tibiotarsus Island’ tibiotarsus King Island emu recent) bones) Undated (probably Pre-1897 femur QVM.1480 Femur Unknown recent) Sandy Cape Vertebra (not Unknown (probably putative emu Unregistered Sandy Cape emu) Holocene or recent) vertebra

vi Contents Acknowledgements iii Abbreviations iii Executive summary iv 1.0 Introduction 2 2.0 Methods and reporting protocol 3 3.0 Inventory of specimens 5 3.1 Scotchtown Cave emu bones (QVM.2005.GFV.45, QVM.2006.GFV.120-121, QVM.2007.GFV.5) 5 3.2 Mowbray Swamp emu bones (QVM.1990.GFV.138-143) 9 3.3 Irishtown tibiotarsus (QVM.1990.GFV.144) 12 3.4 Jackson egg (QVM.1965.GFV.0006) 14 3.5 Mole Creek tibiotarsus (QVM.1991.GFV.54) 18 3.6 Mole Creek (fractured) tibiotarsus (‘QVM.1489’) 20 3.7 Dromaius Cave partial skeleton (QVM.2013.GFV.11, QVM.2016.2.008) 22 3.8 Caveside Emu A partial skeleton (1974.2.5-7, 9) 25 3.9 Caveside Emu B partial skeleton (associated with QVM.1974.2.9) 29 3.10 Mt Cameron West femur (QVM.1993.GFV.146) 31 3.11 Newstead House emu leg (QVM.2002.2.1) 33 3.12 Anomalous ‘King Island’ femur 36 3.13 Anomalous ‘King Island’ tibiotarsus (QVM.1993.GFV.18) 38 3.14 Pre 1897 femur (QVM.1490) 40 3.15 Sandy Cape putative emu vertebra (unregistered) 41 4.0 References 42 Appendix: QVMAG old register 44

1 A significant constraint on the discussion of 1.0 Introduction possible differences in the morphology of The Tasmanian emu (Dromaius novaehollandiae the various emu populations is the dearth diemenensis) is generally considered an island of comparative material from Tasmania. sub-species of the Australian emu (Dromaius By comparison, King Island emu (Dromaius novaehollandiae novaehollandiae). Dudley Le ater) is relatively well represented in public Souef (1856-1923) examined Tasmanian emu collections, chiefly because in the early years skins at the British Museum and then briefed of the 20th century a considerable quantity of the British Ornithologists’ Club, which recorded bones was collected from sandblows on the the following in its bulletin for 1907-08: island. The unfortunate lack of specimens of Tasmanian emu was recognized by Le Souef’s Mr D Le Souef informed the meeting that contemporary, Colonel. WV Legge, who he had recently examined the collection expressed the view that ‘it is most desirable of the skins of Dromaeus in the British that some search be instituted for the bones of Museum, and that among them he had our Tasmanian species’ (Legge 1907). Little or discovered two specimens of the Tasmania nothing seems to have been done at the time. emu, presented to that Museum by the late Ronald Gunn in 1838. These skins were Whereas the extinction of the thylacine of great interest, as the species was now was anticipated by zoos and museums, extinct in Tasmania; and it was evident, which responded by collecting and trading from the specimens in the British Museum, thylacine material, this had no direct parallel that the emu of Tasmania was distinct for Tasmanian emu. Tasmanian emus appear from that of the continent (Dromaeus to have disappeared from the wild by the novae-hollandiae), having no black on the 1860s and it has been suggested that the last throat and fore-neck, these parts being captive bird died in 1873 (Le Souef 1904). entirely white. Mr Le Souef stated that the Apparent similarity of Tasmanian emu and discovery of these facts confirmed the mainland emu may be partly to blame for the opinion which he had already expressed lack of early interest in collecting specimens as to the distinctness of the two species of of the former. Gunn (1852) had commented Dromaeus, based on a study of their eggs; on possibly subtle differences in external and that the name D. diemenensis, which he morphology, but it is possible that the general had proposed for the Tasmanian emu, was perception at that time was that the respective now established by the examination of the emu populations were essentially the same skins above mentioned. animal. An unknown number of mainland emus were imported to Tasmania for use as A century after this thrifty description, the pets or ornamental birds in the early to mid- taxonomic status of the putative Tasmanian 19th century. species/sub-species is unresolved. Although accepted as a valid taxon under the Tasmanian Fortuitously, in the years leading up to and Threatened Species Protection Act 1995, some following the First World War, skeletal bird systematists have reserved their position remains of emus were discovered during on this (Ridpath & Moreau 1966; Christidis drainage works in the Smithton district of & Boles 2008). Recently, DNA evidence north west Tasmania. Some of these were suggests that Tasmanian emu genotype is deposited at QVMAG, thanks to the interest indistinguishable from that of other island and and enthusiasm of Museum director Herbert mainland emu populations (Thomson et al. Hedley Scott. In the 1940s, additional emu 2018). The degree of similarity or otherwise bones were collected by Scott’s son Eric from of the morphology (phenotype) is yet to be Scotchtown Cave near Smithton. Later again, rigorously investigated. QVMAG curator RH Green recovered two partial emu skeletons from a cave at Mole Creek in the 1970s. Over following decades

2 a few other dribs and drabs trickled in, including scanned copies of HH Scott’s including, notably, a third partial emu skeleton acquisitions ledger for the period 1911-1930 discovered by Ross McNeill and Paul Flood in (henceforth referred to as the ‘old register’), a cave at Mole Creek in about 2005. were provided by Natural Science Collection Officer Tammy Gordon. The net result of these episodic acquisitions by QVMAG is that no other public collection In this report, direct transcriptions of in Australia holds a comparable number of handwritten card labels, letters, ledger entries specimens of Tasmanian emu. A review of and similar sources are identified by indented this material was considered timely because italic text or, for short quotes, by enclosing certain specimens had been considered them in inverted commas. Occasional editorial dubiously Tasmanian, due to the possibility insertions and clarifications are provided in that they post-dated the importation of regular text (not italics) in square brackets. mainland birds to Tasmania and might be For example, [blank] means no original entry. mainland escapees or hybrids. This issue was Transcriptions of printed card labels and addressed during this study by radiocarbon headings are provided in regular text with dating relevant bones. The material is original abbreviations and capitalisations. exclusively from northern Tasmania and Measurements quoted in the text were taken Flinders Island (Figure 1). using digital display vernier calipers, where One of the main challenges encountered the measurement is 150 mm or less. These during the project was to verify, as far as results are reported to the nearest 0.1 mm. practicable, the provenance of items collected Measurements in excess of 150 mm (e.g. long over a period of about 170 years, during bones) were made by resting the bone on a which time collecting practices and museum steel rule on a flat surface. A carpenter’s steel protocols have changed profoundly. Whilst it square placed upright against the respective was not possible to verify the provenance of ends of the bone was used to align one end all specimens, some progress was made. It is with the zero and to determine its length on hoped that the results of this work will be a the rule. The results are reported to 0.5 mm, useful foundation for more informed scientific corresponding to the gradations on the rule. research and presentation of Tasmanian emu, Samples for radiocarbon dating of bone were this island’s only representative of the ratite removed using a ‘Dremel’ tool. For larger group of birds. bones, a small portion of bone (~500 mg) Locality data for Tasmanian emu material held was excised using a cutting disk, whereas by QVMAG is plotted in Figure 1. Note: This for smaller bones a hole was drilled into the project did not address QVMAG holdings of bone whilst collecting the resultant powder. King Island emu (Dromaius ater). Eggshell was sampled by breaking off a small portion (~20 mg) after scoring the cuticle with 2.0 Methods and reporting a scalpel blade. protocol Radiocarbon dating was undertaken at the Australian National University radiocarbon The bulk of the work for this project was dating laboratory, Canberra. The dating undertaken at the QVMAG Royal Park procedure for bone (Fallon et al. 2010) precinct over a five-day period in October includes chemical pretreatment to extract and 2017. All confirmed and potential Tasmanian clean collagen according to an ultrafiltration emu specimens and associated card labels protocol.2 Acid is then used to remove the were examined and photographed. Relevant bone mineral and any exogenous carbonates, records from QVMAG specimen databases,

2 Source: http://rses.anu.edu.au/services/anu-radiocarbon-laboratory/laboratory-methods (downloaded 10/7/2018).

3 alkali is used to remove humic substances, and Tasmania. Several specimens were excluded the sample is acidified to dissolve carbonate. on this basis. One sample submitted for dating The sample is then gelatinised by heating in was rejected by the laboratory because of a dilute acid, turning the insoluble collagen insufficient collagen (QVM.2005.GFV.45B). into soluble gelatin allowing large insoluble Radiocarbon dating results received from the particles to be removed with a filter and laboratory were calibrated to correct for global small soluble molecules to be removed with variations in carbon isotopes over geological an ultrafilter. Pretreatment of shell involves time using the University of Oxford’s online removing any soft white recrystallised material software OxCal 4.3.2 (Bronk Ramsay 2017). with a drill and then washing in dilute acid to The calibration curve SHCal13 was applied remove at least 10% of its weight. (Hogg et al. 2013). Uncalibrated radiocarbon On advice from the laboratory, dating was not dates are reported in years before present attempted on specimens coated in varnish (BP). By convention, radiocarbon BP is taken as a museum procedure or on specimens to be 1/1/1950. Calibrated radiocarbon dates which had been immersed for long periods in are reported as a range at 95.4% probability, carbon-rich water (e.g. water-logged peat), meaning that there is a 95.4% probability due to the likelihood of sample contamination that the true age of the specimen falls within by extraneous carbon. This precluded dating the range given. These dates are reported of several important specimens from the as calBC or calAD, which can be read as Smithton area and one from Mole Creek. normal historical years. Radiocarbon dating is Furthermore, dating was considered to serve generally considered unreliable beyond no purpose in the absence of reasonable 50 ka. Dates older than this are typically certainty regarding provenance, such as described as infinite dates. whether the specimen was collected in

Figure 1. Map of Tasmania showing localities of emu fossil collections held by QVMAG.

4 3.0 Inventory of specimens

3.1 Scotchtown Cave emu bones (QVM.2005.GFV.45, QVM.2006. GFV.120-121, QVM.2007.GFV.5)

Plate 1. Scotchtown Cave tibiotarsi (QVM.2005.GFV.45). The radiocarbon date was obtained from the upper tibiotarsus.

Plate 2. Scotchtown Cave tarsometatarsus (QVM.2006.GFV.120).

5 Plate 3. Scotchtown Cave femur fragment (QVM.2006.GFV.121).

Plate 4. Scotchtown Cave phalanges (QVM.2007.GFV.5). The radiocarbon date was obtained from the larger bone.

6 Item(s) removed from towards the distal end of the 2 x partial left tibiotarsi (QVM.2005.GFV.45) shaft on the longer fragment. 1 x partial right tarsometatarsus (QVM.2006. Tarsometatarsus (QVM.2006.GFV.120) GFV.120) 1 x right femur fragment (QVM.2006.GFV.121) The shaft is fractured off cleanly 88 mm 2 x phalanges (QVM.2007.GFV.5) above the distal end (Plate 2). The T2 and T4 trochleae are missing and the remaining Provenance data central trochlea is truncated towards the Card labels and the QVMAG digital database distal end. The bone is whitish in colour indicate that this set of specimens was with a hard, chalky aspect. The surface is collected by EOG Scott from Scotchtown Cave superficially marked in several places, in a near Smithton in 1942. The database notes the manner suggestive of contact with a metal following: tool during excavation. The interior cavity of the bone is filled with dark brown material, Location described by EOG Scott likely a remnant of the cave fill where it was as: “apparently a filled-in cave in a found. Flaky translucent coating on portions of limestone formation opened up by the bone suggests local application of varnish Mr W Archer when quarrying in the or similar product. northern face of a hill (approximately 300 yards long approximately 20 yards Femur (QVM.2006.GFV.121) high) on W Ollington’s property, near This minor fragment is 60 mm in diameter, Smithton, Tasmania Gate of property 3.2 comprising a portion of the articular surface miles southerly from Smithton Council at the proximal end of the bone (Plate 3). Chambers; March 1942.” The colour is yellowish-brown and the bone The discovery of the rich bone bed exposed by lacks the hard, chalky aspect of some other quarrying at Scotchtown Cave, and Scott’s role Scotchtown Cave fossils. Traces of brown soil in salvaging material prior to destruction of the are lodged in the pores of exposed cancellous site by further quarrying, is discussed by Wylie bone; however, the majority of pores are free (2013). of sediment. The bone is coated in varnish or similar product. Description Phalanges (QVM.2007.GFV.5) Tibiotarsi (QVM.2005.GFV.45) The two toe bones are in good condition with The two tibiotarsi fragments are both distal no fractures or significant abrasion (Plate 4). ends of bones fractured across the shafts One of the two is substantially larger than (Plate 1). The specimens are 150 mm and the other (length 55.5 versus 42.0 mm; 129 mm long, respectively. The form of the weight 12.0 versus 5.3 g) and may be the first fractures is spiral in both cases. On the longer knuckle on the large central toe (i.e. DIII P1). fragment the fracture appears somewhat The bones are creamy-whitish in colour with fresh, exposing whitish chalky bone— traces of reddish-brown sediment. The smaller possibly, this fracture occurred during or after phalange appears to have been coated with collection. The fracture on the other bone varnish or similar product; the larger appears bears traces of sediment, implying that it was uncoated. already broken when exhumed from the cave. The distal ends of both bones are abraded to the point whereby most of their articular surfaces are lost. The bones are pale creamy- brown in colour with patchy black staining. Traces of reddish-brown sediment can be observed on both tibiotarsi. Prior to this study, a small rectangular piece of bone had been

7 Dating Four samples were submitted for radiocarbon dating. Three returned results:

Sample Laboratory code Radiocarbon age Calibrated age

QVM.2007.GFV.5 SANU-56324 33 109±681 BP 37 909-33 764 calBC phalange

QVM.2006.GFV.120 SANU-56325 35 793±935 BP 40 131-36 665 calBC tarsometatarsus

QVM.2005.GFV.45 SANU-56326 36 517±1024 BP 40 799-37 138 calBC tibiotarsus

The laboratory report noted that the result Collection Officer Craig Reid, who registered for QVM.2006.GFV.120 (tarsometatarsus) the specimens. should be treated with caution, due to low mass and percentage collagen yield; The three radiocarbon dates on Scotchtown however, the stable isotope and carbon-to- Cave emu material cluster together and their nitrogen ratios were considered adequate. calibrated ranges overlap, indicating that they For QVM.2005.GFV.45 (tibiotarsus), the are essentially contemporaneous. At ~34-41 report noted that this also should be treated ka, these are the oldest direct evidence for with caution. In this case collagen was not emu in Tasmania. The dates may in fact refer ultrafiltered, as low yield was suspected to a single emu, although the presence in during chemistry. Again, despite low mass and the cave of two right tibiotarsi, one of which percentage collagen yield, the stable isotope could not be dated, indicates that at least two and carbon-to-nitrogen ratios were considered emus died there. The dated emu or emus lived adequate. No date was obtained on tibiotarsus during MIS3, at time when global climatic QVM.2005.GFV.45B (the shorter and more conditions were becoming progressively robust of the two tibiotarsi), due to poor colder leading up to the Last Glacial Maximum preservation of bone collagen. (MIS2). Discussion Other bones from Scotchtown Cave are Scotchtown Cave yielded one of the richest potentially older than the radiocarbon results and most diverse Late Quaternary vertebrate reported here. Turney et al. (2008) cite a fossil assemblages in Tasmania. Unfortunately, luminescence date of 56±4 ka for Scotchtown the cave was destroyed and only limited Cave sediment from material at QVMAG. information is available on the nature of the Also, the presence of megafauna at this site bone deposit. A species list is provided by suggests considerable antiquity, as megafauna Gill and Banks (1956), who state that the is not known to have survived in Tasmania cave contained an average depth of two feet beyond about 45 ka BP or shortly after (~0.6 m) of chocolate-coloured cave earth (Turney et al. 2008; Cosgrove et al. 2010; with numerous well-preserved bones. These Gillespie et al. 2012). authors interpret the site as a carnivore den, possibly because three large carnivore taxa were recorded there (see Gill 1953). Their species list for Scotchtown Cave does not record the presence of emu, suggesting that this was not recognised until the collection was examined more recently by QVMAG

8 3.2 Mowbray Swamp emu bones (QVM.1990.GFV.138-143)

Plate 5. Mowbray Swamp emu bones (QVM.1990.GFV.138-143) found by T Edwards and sold to the museum in 1924.

Item(s) THE SMITHTON FIND. 1 x synsacrum 1 x right femur From an old contributor to our 1 x right tibiotarsus palaeontological series of vertebrate remains 2 x tibiotarsi (pair) – Mr. Tom Edwards – there came to us 1 x cervical vertebra in October, 1924, a synsacrum, 1 femur, 1 tibio-tarsus, 2 tarso-metatarsi, and 1 cervical Provenance data vertebra of a Tasmanian Emu recovered from Card labels in HH Scott’s hand, and entries in Mowbray Swamp. These were all associated the old register for acquisition 1488, indicate bones absolutely mature, using the word in that these specimens were purchased from its true osteological sense – as applicable T Edwards in October 1924 (Appendix). The only to bones whose external texture locality is given as Mowbray Swamp near manifest the highest muscular development Smithton. and the super-ossification incidental thereto. Everything considered, and having due The bones are the subject of a brief article by reference to published notes, I consider this Scott (1931), who reported: bird to have been a female. Scott went on to discuss aspects of the morphology of the bones.

9 The specific location and context of the bones tibiotarsus and synsacrum are noticeably darker on discovery is not recorded. It is likely that than the other bones. All are heavily coated in they were unearthed during the construction of varnish as a museum treatment. drains to facilitate farming on hitherto swampy Femur (QVM.1990.GFV.139) country, as per other vertebrate fossils collected in Smithton area in the early decades of the The bone is complete except for moderate 20th century. Their darkly stained condition abrasion at the proximal end and lesser abrasion is comparable with that of these other fossils, and minor cracks at the distal articular surface. A including celebrated skeletons of ‘Nototherium’ small cluster of granular pinkish clasts 2-4 mm in (Zygomaturus), which were found in ditches diameter can be observed lodged in a hollow at excavated into the characteristically peaty the distal end. deposits of the area. Scott had previously purchased fossil material from T Edwards in Tibiotarsus (QVM.1990.GFV.138) 1915 (Scott 1915), noting in the Museum register Minor abrasion can be observed at both ends that these bones were collected from a drain at of the bone. Fine sand was noted falling out of Mowbray Swamp. crevices during handling. A map by Gill and Banks (1956, p. 16) shows Tarsometatarsi (QVM.1990.GFV.140-141) a feature named ‘Edward’s Spring’ off Mella Road, raising the possibility that T Edwards or The two tarsometatarsi are in excellent his family owned the spring and found the emu condition, being complete except for minor fossil nearby. The spring is about 2 km north of abrasion at their proximal ends. Unlike several where Zygomaturus was found on Lovell’s Farm. other tarsometatarsi in the collection, the The presence of the spring raises the further trochleae are intact. possibility that the bones were found in a spring Vertebra (QVM.1990.GFV.142) mound deposit, as such features have yielded well-preserved vertebrate fossils elsewhere in The bone is somewhat abraded and has lost the the Smithton area (Banks et al. 1976; Horton & lateral spine on one side. Murray 1980). Synsacrum (QVM.1990.GFV.143) Additional evidence concerning the stratigraphic The synsacrum has suffered considerable context of the Mowbray Swamp emu is available damage. The fragile ischia and pubis bones are from the presence of sandy materials and tiny missing, as are portions of the dorsal ridge and shell fossils lodged in crevices on the specimens. the proximal and distal ends. Crevices on the This can be observed most clearly on the underside of the bone are occupied by small synsacrum, where a number of shells are lodged mollusc shells up to about 4 mm diameter. within orifices on the ventral side of the bone. The shells have not been examined in detail and Dating no identifications are reported here. However, Smithton area swamp material was not selected it seems likely that these are freshwater snails, for dating due to the risk that it is contaminated corroborating the inference that the fossil was by younger carbon during prolonged immersion preserved in a swamp or spring. The presence in water-logged soil and the more recent of fossils of freshwater molluscs, typically found application of varnish. interbedded with peats, marls and sands, has been reported from other sites around Smithton, Discussion including vertebrate fossil localities (Gill & Banks As discussed above, the Mowbray Swamp emu 1956; Banks et al. 1976; Colhoun et al. 1982). was probably collected from the peaty soil of a former swamp or spring mound. In the Smithton Description area, soils of this type have formed over Last All the bones are darkly stained, presumably due Interglacial age marine sands and are considered to immersion in water-saturated peaty soil (Plate Pleistocene in age (Gill & Banks 1956). Although 5). The colour is mostly beige or brown—the undated, the fossil is therefore almost certainly

10 also of Pleistocene age and, together with the Irishtown tibiotarsus (QVM.1990.GFV.144), may be of greater antiquity than most or all other known Tasmanian emu fossils. Gill and Banks (1956, p. 27) obtained infinite radiocarbon dates (>50,000 ka) for sediments associated with Mowbray Swamp marsupial fossils, as did Banks et al. (1976) for wood associated with a fossil of the large marsupial Palorchestes at Pulbeena. Gillespie et al. (2012, 2015) obtained finite radiocarbon dates on bone collagen from Mowbray swamp fossils, but considered these unreliable due to contamination. Evidence from a combination of pollen, fossil invertebrates and stratigraphy has been used to infer the character of the Late Pleistocene vegetation and environment of the Smithton area in the during the period when the Irishtown emu probably lived (Gill & Banks 1956; Banks et al. 1976; van de Geer et al. 1986; Colhoun et al. 1982).

11 3.3 Irishtown tibiotarsus (QVM.1990.GFV.144)

Plate 6. Irishtown tibiotarsus (QVM.1990.GFV.144).

Item(s) THE EXTINCT TASMANIAN EMU. 1 x left tibiotarsus Of the extinct Tasmanian Emu I have to Provenance data record the finding of a tibio-tarsus, which was recovered from the Pleistocene swamp The tibiotarsus, originally allocated number 1487 at Irish Town, N.W. Tasmania, during some in the QVMAG collection, now QVM.1990. draining operations carried out in 1920. Our GFV.144, is recorded both as Smithton and Museum is indebted to Mr Willes, of this Irishtown in HH Scott’s original handwritten city, and to the finder of the bone—Mr E H card labels. His entry in the old register is silent Fenton—for this interesting specimen, which, on locality (Appendix), whereas a short article from its long immersion in the swamp, published by Scott (1923) gives the locality as must be, beyond all doubt, the leg-bone ‘Irish Town’ [sic]. The current QVMAG database of a Tasmanian Emu. Unfortunately, the lists the locality as Irishtown. There seems little bone is broken at its proximal end, the shaft doubt that the specific locality where the bones terminating 44 mm. below the femoral were collected is at, or near, the small hamlet articular platform. If the amount named be of Irishtown, 8 km south of the regional centre, allowed for, it exactly agrees with a second which is Smithton. similar-sized bone to be dealt with presently. A handwritten display card by Scott states Scott went on to present measurements of the that the bone was found in a swamp by Mr EH bone and to compare these with the Newstead Fenton in 1920 and donated to the Museum House emu leg (QVM.2002.2.1). by Fenton and Mr Willes on 24/8/1920. Scott (1923) later published a brief account of the find: Description The proximal end of the bone is broken and/or eroded away, shortening the overall length by several centimetres—Scott (1923) estimated

12 that about 75 mm of the length was missing. The distal end is fairly intact but also somewhat abraded. The bone on the shaft is cracked superficially and some few hundred microns of outer bone has flaked off over a considerable portion of the surface. The colour is dark brown to beige and noticeably lighter where the surficial bone has exfoliated off. The bone is heavily coated in varnish (Plate 6). Dating Smithton area swamp material was not selected for dating due to the risk of contamination by younger carbon during prolonged immersion in water-logged soil and the more recent application of varnish. Discussion This bone bears comparison to the Mowbray Swamp emu (QVM.1990.GFV.139)­—both specimens are darkly stained by organic compounds suggestive of prolonged immersion in carbon-rich soil water. Their age may be broadly contemporaneous with that of other Mowbray Swamp vertebrate fossils including megafauna (i.e. Late Pleistocene). The probable antiquity of the specimen seems not to have been appreciated at first by curator Scott. His card label states that ‘the bone came from a fully adult bird, and was apparently buried in the swamp for 50 years at least’. As noted above (see Mowbray Swamp emu QVM.1990. GFV.139), more recent estimates of the age of fossil bones from the Smithton area differ from Scott’s initial assessment by three orders of magnitude. Scott (1923) later presented a more informed assessment of the age of the specimen: The point to be noted here is that the bone is beyond all question of Tasmanian origin, since its inclusion into the peaty matrix of the swamp was certainly at a much earlier date than that at which any mainland Emus were imported into Tasmania, and therefore it stands as the earliest known specimen of a Tasmanian tibio-tarsal shaft. [emphasis added] Although presently undated, it is indeed possible that, together with the Mowbray Swamp emu, these are the most ancient fossils of Tasmanian emu collected to date.

13 3.4 Jackson egg (QVM.1965.GFV.0006)

Plate 7. The Jackson egg – three fragments of emu eggshell found on Flinders Island by Frank Jackson in 1917, with original label in HH Scott’s hand.

Items 3 x fragments of eggshell Provenance data The small fragments of eggshell referred to here as the Jackson egg were listed by HH Scott on line 1513 of the old register. That entry and an original card label give the locality as Flinders Island, the finder as Mr Frank Jackson and the date of the acquisition as 11/4/1917 (Appendix). The key reference to the finding of the egg is a succinct handwritten note by Jackson dated 24/3/1917 (Plate 8). The note is held by QVMAG and presumably accompanied the specimen on arrival at the Museum. The locality is specified somewhat precisely as 4 miles (6.4 km) inland and about 600 feet (180 m) above sea level. Jackson stated that he believed the sample to be a portion of emu egg and that he had seen similar fossils at Stokes Point on King Island. The reference to ‘fossil’ assumes particular significance in the light of the radiocarbon result discussed below.

14 Plate 8. Frank Jackson’s letter of 24/3/1917 to the Queen Victoria Museum, Launceston.

Description Dating The Jackson egg comprises three angular A sample of the smallest piece of the eggshell fragments of eggshell (Plate 7). The two larger was submitted for radiocarbon dating and pieces are compact polygonal shards measuring returned a result of 14 458±46 BP (SANU- 17.3 x 13.5 mm and 15.7 x 14.8 mm across 55616). This translates to 15 839-15 421 calBC. their main axes, while the third piece is broadly Discussion triangular and measures 6.7 x 4.6 mm. In total, the sample comprises about two square The original identification of the eggshell as centimeters of shell. The combined weight prior that of emu by Frank Jackson, which HH Scott to sampling was 0.68 g, the two larger being concurred with, is correct based on the texture 0.32 and 0.34 g and the other 0.04 g. The of the outer cuticle, which cannot be confused edges of the fragments do not obviously match with that of any other Australian bird (Beruldsen with each other and apparently derive from non- 1980). Whilst the colour is yellowish-brown contiguous portions of eggshell. rather than the characteristic dark green of emu eggshell, this is consistent with the tendency of The external cuticle displays the diagnostic emu eggs to bleach rapidly (Williams & Vickers- knobbly texture of emu egg. The colour is Rich 1991). Eggshell thickness (1.04-1.11 mm ) pale yellow-brown on the exterior surface and compares closely with the value of 1.1 mm for creamy white on the interior surface. Under low Australian emu cited by Williams (1981). This magnification, the outer cuticle has a bleached considerably exceeds eggshell thickness in the pearly aspect interspersed with irregular patches majority of large birds and is diagnostic for emu of whitish opaque material (Plate 9). The whitish and large extinct ratites (Williams & Vickers- component is interpreted as a secondary mineral Rich 1991). coating rather than an original feature of the egg.

15 Plate 9. Outer cuticle of the Jackson egg under low magnification, illustrating the bleached colouration and traces of presumed mineral coating. The scale is in millimetres.

Scott exhibited the Jackson egg at a reading to eggshell on Flinders Island, stating incorrectly the Royal Society of Tasmania in 1923, during that museum collections held no specimens of which he discussed emu material from King emu from Bass Strait islands other than King Island and the Tasmanian mainland. An account Island. Emu is not listed in The Birds of Flinders of the meeting by The Examiner newspaper Island by Green (1969), although this reference referred to ‘a scrap of egg shell that as far as was focuses on extant species. Sutherland and known was the only material evidence of the Kershaw (1971) acknowledged Jackson’s finding emu having ranged on Flinders Island’ (Anon of emu eggshell on Flinders Island but expressed 1923). Curiously, Scott (1923) did not refer to doubt regarding the provenance. Flinders Island the eggshell in his published version of the is not referenced in an otherwise comprehensive presentation. reviews of fossil emus by Patterson and Rich (1987) or Australian fossil birds by Baird (1991). The eggshell was not acknowledged in the More recently, Hume et al. (2018) referenced the scientific literature until four decades later, Jackson egg as possible evidence for a Flinders when Jeanette Hope mentioned it briefly in Island species or subspecies of emu. her PhD thesis on island biogeography in Bass Strait, dismissing the find as a fragment of an The calibrated radiocarbon date—15 839- ornamental emu egg taken to Flinders Island 15 421 calBC (SANU-55616)—demonstrates by a settler (Hope 1969, p. 232). Hope later conclusively that the Jackson is not recent published a summary paper which did not eggshell and cannot be a fragment of recognise any record of emu on Flinders Island ornamental egg. Rather, it is a Late Pleistocene (Hope 1973). Serventy (1967) was aware that fossil preserved from MIS2 following the Last Frank Jackson claimed to have found emu Glacial Maximum. This was a time of rapid

16 climatic amelioration when Flinders Island became isolated from mainland Australia and Tasmania by rising post-glacial seas. The radiocarbon result resonates with the words of the finder, Frank Jackson, who referred to the specimen as a fossil. Jackson was a keen amateur naturalist and an astute observer who searched for evidence of former Aboriginal occupation on Flinders Island. Whilst we know only that Jackson found the specimen at elevation several kilometres inland of Emita township, it seems possible that he chanced upon it whilst searching the ground for Aboriginal stone tools. Unlike on King Island, emus were not present on Flinders Island or other parts of the Furneaux Group when Europeans arrived there at the turn of the 18th century. Fossil evidence for their presence in the Furneaux Group has previously been recorded from Prime Seal Island, 6 km west of Flinders Island. Here, a stratified human occupation deposit at Mannalargenna Cave was found to contain abundant emu eggshell within layers bracketed by radiocarbon dates of 20.6 and 14.3 ka BP (Brown 1993). Emu eggshell was not detected within the final circa 6 ka of occupation deposit, which ceased about 8 ka BP, suggesting either that emus were no longer taken as food or they had become scarce or extinct at about 14 ka BP. Emu is not recorded from bone-bearing sediment at Ranga Cave on Flinders Island, which samples Furneaux Group fauna of an undefined period prior to about 8 ka BP (Hope 1969), or from the Palana sandblow, which samples it more recently (Hope 1973). In summary, the Jackson egg provides evidence for the presence of Late Pleistocene emu on Flinders Island. The date on this specimen broadly corroborates the age of the youngest dated layers containing emu shell at Mannalargenna Cave on Prime Seal Island, constraining the timing of extinction of emu in the Furneaux Group to circa 14-15 ka BP or later.

17 3.5 Mole Creek tibiotarsus (QVM.1991.GFV.54)

Plate 10. Mole Creek tibiotarsus (QVM.1991.GFV.54).

Item(s) Scott in December 1914. Scott referred to EC 1 x right tibiotarsus Clarke in a published paper of 1931, stating that Clarke had been ‘induced to collect osteological Provenance data specimens from such caves as were immediately The locality for the tibiotarsus, specimen 1489 in available to him’. The paper goes on to state that the old Museum register, is listed as ‘Sassafras Clarke’s collection included ‘almost every animal Farm, Mole Creek’, or simply ‘Mole Creek’ on living in Tasmania to-day’ (sic). Based on their card labels and the current Museum database. different initials, between 1914 and 1931 Scott The finder is listed as EW Clarke and the received bones from two Clarkes at Mole Creek acquisition data as 13/3/1931. The specimen is (EC Clarke and EW Clarke); alternatively, he briefly described in a paper by Scott (1931): may have dealt with one Clarke but on occasion wrongly recorded his second initial. THE MOLE CREEK FIND. From Mr E. W. Clarke, of Mole Creek, we have received a Whereas the Mole Creek locality suggests tibio-tarsus of the Tasmanian Emu. This falls that the tibiotarsus was found in a cave, cave into line with our conceptions of the female bones typically exhibit yellow, beige or reddish bird. It is shorter, but mutilation in the item of colours. In contrast, the dark brown colour of this post-mortem rubbing and grinding accounts bone bears comparison with Mowbray Swamp for about 12 mm., the remainder coming fossils, which are stained by organic-rich soil within the range of individual variation. water. Interestingly, the general colour and size of the Mole Creek (right) tibiotarsus implies that Reference to Sassafras Farm in the old register it, and the Mowbray Swamp (left) tibiotarsus could refer to one of several farms on Sassafras (QVM.1990.GFV.138), could be taken as a Creek, west of Mole Creek township. It is not matching pair. In fact, beyond their superficial stated that the bone was collected in a cave, similarity, there is no evidence to suggest that although pitfall caves are present in the vicinity the two lots of specimens are related. of Sassafras Creek and three other lots of emu bones have been found in caves in the Mole It is possible that the Mole Creek tibiotarsus was Creek area. unearthed during an excavation—for example, construction of a drainage ditch. The country It may be relevant that the old register contains east of Mole Creek was formerly known as the an entry for marsupial bones collected in a cave Western Marshes and extensive drains were at Mole Creek by EC Clarke and received by dug between Ugbrook and Caveside (which

18 is several kilometres from Sassafras Creek). Burial in waterlogged soil could account for the dark colouration; on the other hand, the bone is heavily coated with varnish or similar product, which may have altered or concealed the original colour. It is important to note that QVMAG holds a second emu tibiotarsus attributed to Clarke from Mole Creek in 1931. The second tibiotarsus—referred to in this report as Mole Creek (fractured) tibiotarsus (‘QVM.1489’)—is discussed in the following chapter. Description The bone is abraded at the proximal ends, exposing the spongey interior, but otherwise complete (Plate 10). A significant crack extends along and partly around the shaft at the proximal end. The colour is uniformly yellowish-brown, possibly in part due to the application of varnish, imparting a somewhat sticky feel to the bone when handled. A small plug of what appears to be dark sediment is lodged in a hollow at the distal end. Dating This specimen was not selected for carbon dating because of the potential for contamination from the varnish or similar product with which it has been impregnated. Discussion The tibiotarsus is in good condition and looks ‘old’—it is unfortunate that the specimen is coated and not suitable for radiocarbon dating. The lack of more detailed original documentation is also unfortunate and it is not possible here to provide a more precise definition of when, where and how the bone was collected. Possibly, Scott himself lacked these details, which would account for the fact that he made no record of them.

19 3.6 Mole Creek (fractured) tibiotarsus (‘QVM.1489’)

Plate 11. Mole Creek (fractured) tibiotarsus (‘QVM.1489’) (below), alongside Mole Creek tibiotarsus (QVM.1991.GFV.54) (above). Both bones bear the old registration number ‘1489’ and are attributed to Clarke at Mole Creek in 1931. The intact tibiotarsus is discussed under a separate entry.

Item(s) If, as proposed here, Scott’s entry for 1489 does not refer to the fractured tibiotarsus, then this 1 x partial left tibiotarsus in two pieces bone is of indeterminate provenance. Provenance data Description This specimen bears the number 1489 inked on The bone is fractured into two pieces (Plate 11). the bone, apparently referencing a line entry of The longer portion is 275 mm long, representing that number in the old register. The entry refers one side of the shaft produced by an extended to a tibiotarsus from Mole Creek given to the lengthwise fracture. The nature of the fracture is museum by EW Clarke on 13/3/1931. Similarly, a suggestive of an injury to fresh, green bone, as more recent card label attached to bone states: opposed to older, dry bone. The shorter portion Donated by E.W. Clarke 13.iii.1931 is 134 mm long and comprises the distal portion Transferred from Geology collection c. 2012 of the same bone. The proximal end is missing. where it was found in back of store with other The colour varies from yellowish-white to H.H. Scott material. Recent, not fossil. greyish-brown. A distinct transition from darker to lighter coloration can be seen part way down It is considered very unlikely that the entry for the shaft, in a manner suggestive of staining due 1489 in the old register refers to the fractured to partial burial in, or contact with, hummus or tibiotarsus shown in Plate 11. This is because sediment. A small quantity of fine sandy greyish- the register references a bone in ‘fairly good brown sediment was observed attached to the order [but] rubbed proximally’, as does a interior surface of the fractured shaft. Despite published reference to the specimen by Scott being fractured into two halves, the bone is (1931), who mentioned damage from ‘post- unweathered and appears rather fresh compared mortem rubbing and grinding’. In contrast, this to most other material examined in this study. specimen is missing the proximal end and is Traces of dried soft tissue remain attached at the prominently broken across the shaft into two distal extremity. parts. Additionally, the length of the tibiotarsus cited in Scott’s paper agrees more closely with QVM.1991.GFV.54.

20 Dating This specimen was not selected for carbon dating because the provenance is uncertain. Discussion As noted above, this tibiotarsus appears to have been confused with another bone (Mole Creek tibiotarsus: QVM.1991.GFV.54). If this interpretation is correct, then the fractured tibiotarsus is wrongly recorded on the label and QVMAG database. Available evidence for alternative interpretations of provenance is fairly minimal. The bone is certainly from an emu. It presents as a somewhat stained, incomplete but not obviously mineralised or weathered item suggestive of a found object exposed for a period in the environment. It is not a professionally prepared osteological specimen; it might be from a cave but nothing obviously confirms this. The fact that the bone retains traces of soft tissue implies that it cannot be of any great antiquity, unless preservation conditions were exceptional, which is possible under certain conditions (e.g. desiccated thylacine carcasses found in caves on the Nullarbor Plain). Unfortunately, these observations and inferences provide no real clarity regarding the actual origin and history of this enigmatic specimen.

21 3.7 Dromaius Cave partial skeleton (QVM.2013.GFV.11, QVM.2016.2.008)

Plate 12. Dromaius Cave tibiotarsus (QVM.2013.GFV.11). This bone was collected and registered in 2013, prior to collection of the remainder of the skeleton in 2016.

Plate 13. Dromaius Cave emu bones (GFV.2016.2.2008). From left to right, the unbagged long bones are: tarsometatarsus, fibula, tibiotarsus and femora. The bagged bones include phalanges, fragments of ribs, vertebrae and a minor piece of the skull. The collection includes two bones (not shown here) stored separately in inert gas, to ensure optimal preservation. Item(s) bones are: QVM.2013.GFV.11 2 x femora 1 x left tibiotarsus in 4 pieces 1 x right tibiotarsus 1 x unidentified bone fragment (not emu) 2 x fibulae QVM.2016.2.008 2 x tarsometatarsi This collection comprises approximately 70 10 x phalanges pieces of bone, representing roughly half this 2 x cervical vertebrae number of individual bones. The more complete ~9 x sternal ribs

22 ?synsacrum (fragments) DNA analysis, the bones were transferred to ?scapulocoracoid (fragments) QVMAG for cleaning and drying. Skull (fragment of maxilla) Description QVMAG also holds a quantity of marsupial QVM.2013.GFV.11 bones, notably Macropus rufogriseus and Thylogale billardierii, collected during the The proximal end of the tibiotarsus is fragmented excavation of QVM.2016.2.008 in 2016; also, and incomplete. The remainder is broken into samples of sediment, charcoal and gravelly two roughly equal portions by a transverse residue retained after washing the excavated fracture across the shaft. Much of the bone is spoil to extract smaller bone content. coated in light brown clayey silt from the cave. The colour is otherwise light creamy-yellow. Provenance data The bone feels light and potentially brittle. The provenance of this most recent Tasmanian An unidentified bone fragment 69 mm long, emu acquisition by QVMAG is not in question. probably macropod long bone, is held under the The bones were found in a cave at Mayberry same registration number. near Mole Creek by Ross McNeill and Paul QVM.2016.2.008 Flood in about 2005. Both men were then employed by the Parks and Wildlife Service at The bones are variously fractured and abraded. the Marakoopa Cave field centre and, in their For example, the proximal ends of both femora free time, found and explored caves, including are missing; their distal ends have survived caves on land owned at that time by Deidre but in fairly abraded condition. The left Smith. Within one of these caves, McNeill and tarsometatarsus has lost the central trochlea, Flood noted the presence of numerous bones, whereas the right tibiotarsus has lost the a small sample of which they collected and proximal end via a sharp spiral fracture across gave to Smith, who forwarded it to QVMAG in the upper part of the shaft. Many of the smaller February 2005. The larger bone in the sample bones are broken or incomplete. was recognised as emu by Craig Reid (QVMAG), who registered it as QVM.2013.GFV.11. Traces of brown clay silt can be observed adhering to the bones in some cases; however, McNeill and Flood reported the discovery of the the majority of sediment was removed when the bones to the (then) Nature Conservation Branch bones were cleaned at QVMAG. Three bones of the Department of Primary Industry, Parks, (femur, tarsometatarsus, phalanges) are stored Water & Environment (DPIPWE). A follow-up separately in argon gas to assist preservation. investigation by the Parks and Wildlife Service The bones vary in colour from pale orange to and Natural Values Conservation Branch of creamy-yellow. They are light and somewhat DPIPWE confirmed that the cave contained chalky in texture; there is no obvious evidence of a rich repository of pitfall bones. Unaware mineralisation of their fabric. that McNeill and Flood were using the name Cemetery Cave, Eberhard (2014) applied the The left tibiotarsus and right femur were name Dromaius Cave, referencing the emu sampled for DNA and radiocarbon dating. bones. Dating The land containing the cave was subsequently The right femur (field no. DC/2016/1) returned purchased by the Crown as an addition to the a radiocarbon date of 1 010±30 BP (SANU- Mole Creek Karst National Park. Over three days 49415). Calibrated, this becomes 1 021-1 152 in April-May 2016, the author, supported by calAD. David Maynard (QVMAG), David Thurrowgood (QVMAG) and Kieren Mitchell (Australian Centre for Ancient DNA, University of Adelaide), excavated the remainder of the emu skeleton. Following collection and sampling on site for

23 Discussion scattered nearby at the surface or buried in silty Dromaius Cave comprises a narrow entrance sediment to a depth of 10 cm (Plate 14). in the base of a low limestone cliff. Such is the The emu fossil comprises a partial skeleton in narrowness of the surface opening that it is about 70 pieces, including the major bones difficult to envisage that it could entrap an emu. of both legs in substantially intact condition. However, it can be assumed that the entrance Significant portions of the skeleton are missing, was formerly more spacious, prior to partial such as the majority of the skull (represented blockage by fallen boulders. This possibility is only by a small fragment of maxilla), most of the corroborated by the fact that recovery of the vertebrae and the synsacrum. The provenance emu bones was delayed when a tree fell across of the bones is not in doubt and the particulars the cave, dislodging sizeable boulders from of their finding and recovery are now on record the slope above. It was necessary to remove (this report; see also Eberhard 2014). These boulders that had fallen onto the entrance before facts underpinned selection of this specimen work in the cave could continue. for analysis of Tasmanian emu DNA. The DNA Beyond the constriction at the entrance and analysis, as reported by Thomson et al. (2018), the first few metres beyond, the cave falls found that the Tasmanian emu genotype differs away vertically and becomes progressively little from that of the extant Australian mainland more open, shortly intersecting a high rift 1-2 m populations and the extinct populations of King wide. The base of the rift is a flattish floor at a Island and Flinders Island. depth of about 10 m below the entrance. This The Dromaius Cave emu is one of three partial portion of the cave is episodically inundated by emu skeletons recovered from caves at Mole groundwater, which invades the cave from lower Creek. It is a singular fact that all three emus levels in the karst system after heavy rain. The returned similar radiocarbon dates: 1021-1152 emu bones were found at the base of the rift calAD (Dromaius Cave), 1043-1223 calAD close to the fall line below the cave entrance, (Caveside Emu A) and 895-1021 calAD together with the bones of macropods. The (Caveside Emu B). This clustering of fossils at site represents a classic cave pitfall, containing around 900-1200 AD is at least suggestive of the remains of animals which have perished increased abundance of emus in the Mole Creek after falling into the entrance and becoming area during this period. entrapped. Several of the emu leg bones were found in semi-articulated condition; others were

Plate 14. Dromaius Cave emu bones prior to excavation in April 2016. F: femur; Tmt: tarsometatarsus; Tot: tibiotarsus; (l): left; (r): right.

24 3.8 Caveside Emu A partial skeleton (1974.2.5-7, 9)

Plate 15. Caveside Emu A partial skeleton. The synsacrum (SY), femur (FE) and ischia (IS) are registered as QVM.1974.2.5, QVM.1974.2.6 and QVM.1974.2.7 respectively; the remainder of the bones are registered as QVM.1974.2.9.

In this report ‘Emu A’ refers to the larger of two partial emu skeletons recovered from a cave in the Caveside-Mole Creek area by R Green and A Alexander in 1972-74. This material was registered by Green in four lots: QVM.1974.2.5, QVM.1974.2.6, QVM.1974.2.7 and QVM.1974.2.9. Green did not register Emu B, which is discussed under a separate entry (see Caveside Emu B below). The two emus are referred to here as ‘Caveside emus’, in order to reduce scope for confusion with other emu material collected from the Mole Creek area.

25 Plate 16. View of the cranium and mandible of Caveside Emu A (part of QVM.1974.2.9).

Item(s) Provenance data Emu A is represented by 70 bones registered in The card labels and database entries against four lots. the four registration numbers on the specimens indicate that the synsacrum (QVM.1974.2.5), QVM.1974.2.5 femur (QVM.1974.2.6) and detached ischia 1 x synsacrum (QVM.1974.2.7) were collected by A Alexander QVM.1974.2.6 from a cave near Mole Creek in 1972. The 1 x left femur remainder of the skeleton, including the skull (QVM.1974.2.9), was collected by QVMAG QVM.1974.2.7 curator RH Green from ‘a small limestone 2 x ischia cave’ between Mole Creek and Caveside on QVM.1974.2.9 11/3/1974. There is no reason to believe that 1 x cranium (two pieces) all these items were not collected from the 1 x mandible (two pieces) one cave, together with additional unregistered 22 x vertebrae material discussed under the entry for Caveside 2 x fibulae Emu B. 2 x scapulacoracoid The following additional details are recorded 2 x humeri against QVM.1974.2.9 on the QVMAG digital 2 x tibiotarsi database: 2 x tarsometatarsi 14 x ribs Remains first discovered by Mr Arthur 7 x false ribs Alexander of Stoodley, 1972. From 13 x phalanges newspaper cutting “The bird had apparently fallen into the cave through a small hole in

26 the roof and died of starvation when unable parts of several others display patches of whitish to escape.” deposit, possibly calcium carbonate deposited by cave drips. Traces of brown soil can be observed The newspaper cutting referenced above has not lodged within pores at the proximal ends of the been sighted. tarsometatarsi. Description Dating Synsacrum (QVM.1974.2.5) A sample of bone from the right tibiotarsus The body of the synsacrum is largely intact with returned a radiocarbon date of 927±37 BP minor abrasion at the distal end exposing the (SANU-56323), which translates to 1043-1223 spongey interior tissue. The slender ischia and calAD. left pubis are broken off. The bone is somewhat Discussion weathered on the left side, which is affected by surficial flaking across about half of the surface. The collection history of the Caveside emus, as The right side is virtually unweathered. The gleaned from RH Green’s card entries, can be ventral surface shows traces of whitish mineral summarised as follows: deposit, possibly calcium carbonate deposited by • 1972. Mr A Alexander of Stoodley (18 km cave drips. The bone is creamy white in colour. north of Mole Creek township) collected Lack of staining or sediment lodged in crevices several large bones found by him in a cave indicates that the bone has not been buried. between Mole Creek and Caveside. Femur (QVM.1974.2.6) • 7/2/1974 Alexander forwarded the bones Some abrasion can be observed on the upper to QVM. side at the distal end, exposing cancellous bone. • 11/3/1974 Green returned to the cave and The bone is otherwise virtually unweathered, collected the additional bones, including intact and remarkably ‘fresh’ in general aspect. an emu skull and portions of the skeleton of a The colour is pale creamy white, with some second emu. evidence of light sediment staining on the central and distal portions of shaft. The various registration numbers imply that Green registered the initial donations by Ischia (QVM.1974.2.7) Alexander in three lots (QVM.1974.2.5-7), These two fragments of the synsacrum are in prior to himself visiting the cave and recovering similar condition to the main portion of it (see additional bones two years later. It can QVM.1974.2.5 above). The reference to ‘ribs’ on reasonably be assumed that when all the bones the card entry is incorrect. were in front of him, Green recognised that the collection represented two emus, one larger and Skull and post cranial bones (QVM.1974.2.9) more complete than the other. He registered The majority of bones in this lot are intact and the bones of the larger animal in one lot in good condition. Fractured and/or incomplete (QVM.1974.2.9). As he had already separately bones include the right side scapulacoracoid, registered the bones previously collected several ribs—three have been glued—and the by Alexander, also of the larger bird, these skull, which is incomplete. The tibiotarsi and retained their separate registration numbers tarsometatarsi are in almost perfect condition. (QVM.1974.2.5-7). Database entries for these The vertebrae are somewhat worn and several initial specimens state that they were originally have lost parts of their spinous portions. The incorrectly entered as Cape Barren goose, bones vary in colour from whitish or creamy to (Cereopsis novaehollandiae). Possibly, the database pale brown. The right tibiotarsus is affected by a error occurred because at one time emu material slight greenish tinge on the middle portion of the was stored in a drawer adjacent to Cape Barren shaft, in a manner suggestive of moss or algal goose material (T Gordon, pers. comm.). growth. The opposite side of the same bone and

27 Green’s rationale for not registering the bones are not countenanced in a printed card which of the smaller emu is noted on the card entry: may at one time have been displayed with the ‘I have not registered these [other emu bones] specimen. The card states: because of their fragmented nature + confusion THESE BONES ARE PART OF THE with the other bird.’ In other words, it seems SKELETON OF AN EMU. THEY WERE he decided not to register this portion of the FOUND RECENTLY IN A CAVE NEAR collection because the damaged condition MOLE CREEK, TASMANIA. of some bones meant that they could not be confidently allocated to either the larger or the EMUS WERE ONCE COMMON IN smaller bird. In fact, a proportion, including the TASMANIA BUT DIED OUT HERE OVER femur, tibiotarsus and multiple vertebrae, clearly 100 YEARS AGO. belong to the second smaller bird. It is unclear why Green choose not to register at least some VERY LITTLE TASMANIAN EMU of these bones as a separate lot. MATERIAL HAS SURVIVED AND SUCH FINDINGS ARE OF SCIENTIFIC The presence of multiple emu skeletons in IMPORTANCE AS THEY HELP TO BUILD a single Tasmanian cave is unusual but not UPON OUR KNOWLEDGE OF THE BIRD. without precedent (e.g. Scotchtown Cave). Importantly, Emu A is more complete and in better condition than any other known Tasmania emu fossil. The survival of the fragile skull is especially noteworthy. Only two other portions of Tasmanian emu skull are held in Tasmanian public collections and both are minor fragments only—e.g. the tip of a mandible collected at Dromaius Cave (QVM.2016.2.8) and a minor fragment in the TMAG collection. The context of the bones in the cave is not known. For example, were the bones scattered on the cave floor or found in partly articulated condition? Were some of them buried in sediment and if so what was its nature? Were the two emu skeletons close together or separate? Such details would provide useful insights into the taphonomic history of the fossils. Locality information is also sketchy, other than that the bones were found in a small limestone cave of the pitfall type between Mole Creek and Caveside, an area which contains many possible candidate pitfalls. The radiocarbon result indicates that both Emu A and Emu B died hundreds of years prior to European settlement of Tasmania. Emu A is the younger of the two by between about 58 and 338 years, based on the calibrated age ranges. The radiocarbon results dispel any doubt that these emus are genuine Tasmanian birds, as opposed to imported mainland escapees or hybrid descendants, as has been suggested might be the case. In fact, these possibilities

28 3.9 Caveside Emu B partial skeleton (associated with QVM.1974.2.9)

Plate 17. Caveside Emu B post cranial bones (associated with QVM.1974.2.9). Approximately 50 smaller fragments are not shown in this image. The long bone (tibiotarsus) closest to the scale was sampled for dating.

In this report ‘Emu B’ refers to the smaller of is associated with the registered Caveside Emu A two partial emu skeletons recovered from a (QVM.1974.2.9) collected by RH Green from ‘a cave in the Caveside-Mole Creek area by RH small limestone cave’ between Mole Creek and Green (QVM) and A Alexander in 1972-74. The Caveside on 11/3/1974. One of two otherwise larger, more complete skeleton of Emu A was virtually identical card labels for QVM.1974.2.9 registered by Green in four lots: QVM.1974.2.5, includes a note by Green on the reverse side. QVM.1974.2.6, 1974.2.7 and 1974.2.9. Green did This refers to the unregistered bones: not register Emu B. These bones were collected together with Item(s) QVM.1974.2.9 + some belong to that Emu. This specimen comprises approximately 90 However, others including the femur, tarsus bones and bone fragments: + vertebrae are from another Emu, smaller, + probably a female. I have not registered 1 x synsacrum (fragment) these because of their fragmented nature + 1 x sternum (fragment) confusion with the other bird. 17 x vertebrae 1 x right femur For further details see the entry for Caveside 1 x right tibiotarsus Emu A (QVM.1974.2.5-7, 9). 2 x radii Description 2 x ulnae The bones are moderately weathered and porous 7 x ribs (plus fragments) (Plate 16). Very few bones are whole and many false ribs (fragments) are reduced to minor fragments. The more minor fragments (circa 50) intact pieces include vertebrae (especially the Provenance data small caudal elements) and the femur, although The Caveside Emu B is not registered, although it portions of both ends are worn away or otherwise lost. The tibiotarsus is represented only by the

29 shaft of the bone with the extremities gone. the context of the bones in the cave are lacking and, indeed, the cave where they were collected The colour of the bones is pale brown. Some is not known. bear traces of brown soil and/or whitish mineral precipitate, potentially calcium carbonate The radiocarbon result indicates that Emu B deposited by cave dripwater. Close observation died nearly a thousand years prior to European of the femur and tibiotarsus reveal patterns of settlement of Tasmania. It is younger than Emu shallow transverse scratches up to 2 cm long. A by decades to centuries. At 1021 calAD, the These features are discoloured with sediment upper limit of the calibrated age aligns with the and cannot be attributed to damage from lower limit of the calibrated age of the Dromaius handling during or after collection. A circular Cave emu (GFV.2016.2.2008), also from the depression fracture a few millimetres in diameter Mole Creek area. at the distal end of the femur is suggestive of a puncture from an animal bite mark. The tibiotarsus had been previously cut across the shaft with a mechanical tool and was re-sampled at this point for radiocarbon dating during this study. Interestingly, the fragments include a small piece of what may be the horny external structure of the beak. Dating A sample of bone from the tibiotarsus returned a radiocarbon date of 1,120±22 BP (ANU-52431). This result intersects the SHCal13 calibration curve at two points, implying a 26.4% probability that the true age of the sample lies between 895 and 934 calAD and a 69.0% probability that it lies between 958 and 1021 calAD. The laboratory report states that the carbon to nitrogen ratio and stable isotope results do not suggest any major source of contamination in the sample, but the date should be viewed with less confidence than normal because the sample weight was low and yielded a small percentage of dateable material. If the date is inaccurate, it is most likely to be too young. Discussion This intriguing specimen represents a second animal (Emu B) collected from the same cave as the virtually complete skeleton of another, larger bird (Emu A: QVM.1974.2.5-7, 9). The card label by Green states that Emu B is probably female. In fact, most ratite birds including emus display pronounced reversed sexual size dimorphism (i.e. males small, females large). Compared to Emu A, the bones are less complete, more weathered and darker in colour. Unfortunately, as discussed under Emu A, key details regarding

30 3.10 Mt Cameron West femur (QVM.1993.GFV.146)

Plate 18. The very weathered Mt Cameron West emu femur (QVM.1993.GFV.146).

Item(s) The cancellous interior has been largely stripped 1 x right femur, shaft away. The specimen is a pale, creamy colour and looks bleached and brittle. Minor traces of Provenance data an unknown reddish-orange coating are present The relevant labels and database entry provide towards the distal end. minimal details for this specimen, which is recorded only as collected at Mt Cameron Dating West in north west Tasmania in 1935. There The specimen was not selected for dating is no reason to doubt that the bone is genuine because sampling would cause significant Tasmanian emu, beyond the remote possibility damage to an already fragile specimen. that it is derived from a stray imported mainland bird, or the nebulous concern that the Tasmanian Discussion emu population interbred with imported A locality and date is attached to this specimen mainland birds during the 19th century. Wild (Mt Cameron West, 1935); however, the finder emus were not known to be present in north and circumstances of discovery are unknown. west Tasmania in historical time; however, The bleached, weathered condition of the bone imported captive birds were kept at Circular suggests prolonged exposure to the elements, Head in the 1850s (Gunn 1852, Dickson 1926). potentially consistent with the condition of bones collected from coastal sandblows. Sandblows Description commonly contain significant concentrations The bone is extremely weathered and comprises of bone in the form of lag deposits following approximately 50% of the distal portion of the deflation of formerly overlying sand. The shaft (length 114 mm; weight 106.8 g). The majority of King Island emu material in public bone generally is pitted and locally paper-thin. collections was obtained from such features.

31 Coastal dunes around Mt Cameron West are affected by several large sandblows. If the femur was indeed collected from a sandblow then this does little to constrain its age. Dunes on other parts of the Tarkine coast have returned thermoluminescence dates of Holocene to Late Pleistocene age—i.e. hundreds to tens of thousands of years (McIntosh et al. 2009)— implying that the femur could be equally old or young. This, and the imperfect condition of the bone, limits its value for morphological analysis and most other purposes. However, the specimen extends the known range of Tasmania emu to the far north west of the island.

32 3.11 Newstead House emu leg (QVM.2002.2.1)

Plate 19. Newstead House emu leg (QVM.2002.2.1).

Plate 20. Detail of the foot of the Newstead House emu leg. Note the ragged aspect of the tissue at the missing toe.

Item(s) to 9/2/1921 and states that the specimen was 1 x lower right leg (tarsometatarsus) and foot of ‘obtained from the late Ronold [sic] Gunn from St emu; bone and soft tissue in dried condition. Pauls Plains (bird hatched from egg by a Turkey hen)’ (Appendix). Various associated card labels Provenance data repeat or add to these basic details, including This specimen is listed on line 1486 of the old a handwritten display card by HH Scott, which Museum register, which dates the acquisition states:

33 Part of the leg of the now extinct Tasmanian 1852; Scott 1923; Vickers-Rich & Rich 1993, Emu. This specimen once formed part of the Hume 2017). It would appear that this is the collection of the late Ronald Gunn Esq, + is of only surviving verified soft tissue specimen of special interest as its study called out the first Tasmanian emu in Australia. An emu leg of suggestion that the Emu of this Island was unknown provenance, beyond the fact that in distinct in species from the Australian form. the early 1920s it was part of William Ratcliffe’s Gunn’s note that that effect, was published private museum at Port Arthur, is held by the in the Proceedings of the Royal Society of Port Arthur Historic Site Authority. Two or more Tasmania for 1852 (Page 170). Given to the complete skins of Tasmanian emu are held in Museum by the Trustees of the Gunn estate. European collections (Dooley 2017). (NO-1486). Given the importance of this specimen, all Gunn himself corroborated elements of this details of its history are of interest and warrant account in a letter to the Royal Society in 1851, scrutiny. Scott’s entry in the old register states which published this extract: that the Museum acquired the specimen from Newstead House, former residence of Ronold A leg of a Tasmanian emu is now in my [sic] Gunn, in February 1921. Whilst this sounds possession, and so far as I can judge from straightforward enough, Gunn died in 1881 and it, as a very imperfect specimen, there is not known to have referenced the leg since are differences in the arrangement and mentioning it in a letter to the Royal Society of size of the scales, which may justify the Tasmania dated 17/11/1851, an extract of which separation of the Tasmanian emu from that was published by the Society in 1852. Scott of New Holland. Still farther research and himself noted that the leg had disappeared into examination are necessary. obscurity for many years, stating that ‘after Aspects of the provenance of this specimen are resting in the cellar of Newstead House for some explored further in the discussion section below. 70 years, this Emu’s leg has now come to light again, and is upon the table before me as I write’ Description (Scott 1931). The specimen is 522 mm long from the head of the tarsometatarsus to the end of the It is entirely possible, and indeed likely, that extended central toe. It is dried ‘flat’ with the the emu leg referred to in Gunn’s letter and toes depressed downwards and aligned with the that given to the Museum by the trustees tarsometatarsus bone. In life, this posture would of his estate in 1921 are one and the same. be unnatural for a standing emu. The inside toe On the other hand, no original label by Gunn is missing below the first knuckle and the soft has survived and the specimen’s chain of tissue at the injury has a ragged, torn aspect custody over 70 years is assumed rather than with no obvious indication of healing. The outer demonstrated. Gunn’s interest in emus was no scaly layer of skin has largely been stripped off secret and it is plausible that he collected and around the tarsometatarsus. The skin generally stored other emu material at Newstead House. is shrivelled and cracked, exposing underlying He is known to have sent a pair of emus to tissue and bone. Joseph Hooker in London in 1837 and at one time kept live emus on his property, including Dating one of Victorian stock which died by drowning This specimen was not selected for dating in a river (Gunn 1852). On balance, however, because it is recent material. Scott’s conclusion is reasonable: this is probably the leg which Gunn wrote about in 1851. No Discussion evidence to the contrary has come to the This celebrated emu leg is figured in the book author’s attention. Treasures of the Queen Victoria Museum and Other details which deserve scrutiny include Art Gallery (QVMAG 2006, p. 101). It is also Scott’s assertion that Gunn obtained the leg referenced in several other publications (Gunn from St Pauls Plains, where it had been hatched

34 from an egg placed under a turkey hen. Scott these birds reputedly survived until 1873 when it recorded this information in the old register and drowned trying to cross a flooded river. Le Souef later published it in the Papers and Proceedings of recorded that Ranson believed the death of this the Royal Society of Tasmania (Scott 1923). Scott’s bird signalled the extinction of Tasmania emu. source appears to be J Milligan, Secretary of the Finally, a further noteworthy aspect of the Royal Society. Milligan inserted the following specimen is its poor condition. The skin and footnote in the Papers and Proceedings on the underlying tissue is shrunken and fissured by page where the extract of Gunn’s 1851 letter was cracks exposing the underlying bone. Scott published: (1923) commented on this, observing that ‘the Captain Hepburn, of St. Paul’s Plains, outer cuticle has peeled off the scutes [scales]’. possesses a breed of Tasmanian emus, Scott alluded also to the partial loss of one of which he succeeded in rearing from eggs the toes. The present investigation observed found many years since upon the high that skin at the injury on the amputated toe has heathy land in his vicinity. Mr. J. Hepburn a ragged, torn aspect (Plate 20). A mechanism informs me that the booming noise is not of injury involving significant force or torsion, as peculiar to the female, and that the male bird opposed to cutting with a sharp implement or does, though not frequently, make the same gnawing by animals, is inferred. The injury shows sound. The Tasmanian emus share the toils no evidence of healing and seems to be post- of incubation between the sexes, but upon mortem damage. Overall, the condition of the the mother devolves the care of bringing up specimen is rather degraded, and it has clearly the young brood, to which the male parent, suffered serious neglect or rough handling. for the most part, displays an unnatural and The fact that Gunn (1852) referred to it as ‘a most bitter antipathy. very imperfect specimen’ implies that some or all of the damage occurred prior to it coming Milligan’s footnote does not state that Gunn’s into his possession. Decades in the cellar at emu leg came from Hepburn’s flock, although Newstead House may have compounded earlier this would be an obvious thing to record if it deterioration. was known to him. Gunn’s extract does not state that his emu leg came from a bird raised In summary, the Newstead House emu leg by Hepburn. Therefore, we must consider the is important as one of very few soft tissue possibility that Scott jumped to the conclusion specimens of Tasmanian emu in existence, albeit that Milligan’s footnote—plausibly, an interesting only a small portion of the total animal. More aside on emus intended to complement Gunn’s complete soft tissue specimens are available note, not an explanatory addition to it—provided in Europe (e.g. skins at the Natural History critical data on the provenance of the Newstead Museum, Tring, UK); however, this is the only House emu leg. Scott’s close reading of Gunn confirmed example in Australia. and Milligan is confirmed by the fact that he quoted them in full in his own account of the Newstead House leg published in 1923. Scott’s article references no other primary or secondary documentation. Further investigation of archival sources which could shed light on this issue is to be encouraged. An interesting aside on the above is provided by Le Souef (1904), who corroborated Hepburn as a finder of emu eggs. According to Le Souef’s source, Mr Ranson of Killymoon near Fingal, Hepburn raised two generations of captive Tasmanian birds from a clutch of eight or nine eggs found by him in an emu’s nest. One of

35 3.12 Anomalous ‘King Island’ femur

Plate 21. Anomalous ‘King Island’ femur (centre) compared with femora of mainland emu (left: Melbourne Museum B12822) and King Island emu (right: QVM.1971.39.59).

Item(s) Dating 1 x right femur This specimen was not selected for dating because of its uncertain provenance. Provenance data This specimen was found stored with King Discussion Island emu femora, most or all of which are The femur came to attention because of its marked with folio or accession numbers or anomalous size compared to the smaller King collection details directly on the bone. In Island femora with which it is stored, raising the contrast, the femur is not marked in any way and possibility that the bone may be from Tasmanian is not obviously associated with any particular emu. Certainly, the robust size of the specimen, card label, although a number of unattached which is 229 mm long, cannot be reconciled handwritten labels from the early 19th century with King Island ‘dwarf emu’, which were very were noted as stored in the same drawer. much smaller than mainland or Tasmanian emu. Spencer and Kershaw (1910) measured 64 King Description Island emu femora, none of which exceeded 186 This femur is in excellent condition with the mm in length and the majority did not exceed exception of minor abrasion at the extremities 170 mm. (Plate 21). A small hole has been drilled transversely across the bone at the distal end. It has been suggested that Tasmanian emu The colour is pale yellow with a slight waxy were close in size to mainland emu or possibly sheen. There is no evidence of mineralisation or slightly smaller, although this remains to be staining. Close inspection reveals traces of fine verified. The largest emu femur confirmed as whitish powder adhering to the surface. Tasmanian in the QVMAG collection measures 229.5 mm (Caveside Emu A: QVM.1974.2.9).

36 This exceeds femora length in all mainland emus sampled by Patterson and Rich (1987), who recorded a maximum length of 218 mm (n=22). Thus, based on size, the femur in question could be a very large Tasmanian or mainland emu. Its presence amongst the King Island material suggests that it was left there after being used for comparative purposes (see also QVM.1993. GFV.18). Furthermore, the condition of the bone corroborates the possibility of it being a professionally prepared osteological specimen rather than a found object.

37 3.13 Anomalous ‘King Island’ tibiotarsus (QVM.1993.GFV.18)

Plate 22. Anomalous ‘King Island’ tibiotarsus (centre. QVM.1993.GFV.18) with tibiotarsi of mainland emu (top: Melbourne Museum B12822) and King Island emu (below: QVM unregistered).

Item(s) the aforementioned eight King Island femora. 1 x left tibiotarsus The current database entry for QVM.1993. 1 x left fibula (attached) GFV.18 repeats the suggestion that the tibiotarsus is from King Island, while also cross Provenance data referencing the old register and footnoting The provenance of this specimen is highly ‘Marawah Swamp’ as a locality, probably confused. Two card labels associated with the meaning Mowbray Swamp. bone refer to ‘Dromaius minor from King Island’ Description while a third, a scrap of paper, states ‘LEG BONE – TASMANIAN EMU?’. A further tag printed The specimen is large (length: 446 mm) with ‘1502’ seems to link the specimen with a line no indications of weathering. Traces of dried soft entry in the old Museum register, which refers tissue remain attached to the shaft and hold the to eight King Island tibiotarsi received from fibula in life position (Plate 22). The colour is James Mackie Bowling in 1905-07. That entry mostly yellow but darker in places due to traces notes that the King Island bones range in length of fat and dried tissue. from 245 to 356 mm, whereas the tibiotarsus Dating in question is 446 mm long, much larger than This specimen was not selected for dating any King Island emu on record and larger than because of its uncertain provenance. average for mainland emu (cf. Patterson & Rich 1987). Discussion Compounding the uncertainty, the following As noted above, the labelling of this specimen is are written directly on the bone: ‘?1487’, confused and contradictory. The two old register ‘?1502’ and ‘QVM:1993:GFV:18’. From the old folio numbers associated with the bone are register, Scott allocated 1487 to the Irishtown impossible to reconcile with the specimen itself, tibiotarsus (QVM.1990.GFV.144) and 1502 to which is neither King Island emu nor Smithton

38 swamp fossil. It is inferred that the bone is probably mainland emu or, less likely, Tasmanian emu, which has become mixed in with King Island emu material and labelled incorrectly. This situation has parallels with a large emu femur stored with King Island emu femora (see Anomalous ‘King Island’ femur).

39 3.14 Pre-1897 femur (QVM.1490)

Plate 23. Pre-1897 femur (QVM.1490).

Item(s) otherwise largely intact (Plate 23). The shaft is 1 x left femur marked by several fine longitudinal fractures. The colour is pale grey with no obvious staining from Provenance data sediment or organics. The bone is varnished. HH Scott’s entry against 1490 in the old register Patchy darker colouration, most obviously at the identifies the specimen as Tasmanian emu distal end, is suggestive of traces of fatty tissue followed by a question mark (Appendix). It also or perhaps coagulated varnish. states: Dating No very exact history – I found [this femur] This specimen was not selected for dating in 1897 in the old museum, listed as a because of its uncertain provenance and “Deers’ bone” to contrast with some New potential for contamination from the varnish Zealand Moa remains. It may or may not coating. be Tasmanian - it is sightly immature in the sense of being devoid of all super ossification. Discussion I presumed that Alex Morton gave the bone Scott’s entry in the old Museum register or acquired it prior to 1897 suggests that this specimen was displayed at QVM in 1897 as an example of a deer bone! A more recent note on card held with the Whilst it is clearly emu, not deer, very little can specimen records an examination of the bone be said regarding provenance, other than that by a visiting expert in 1971, who evidently this specimen was part of an older collection and considered it likely to be mainland emu and its origins are uncertain. At this stage it cannot unlikely to be King Island emu. Reference on the be confirmed that the femur is either Tasmanian card to ‘No. 1480’ appears to be a transcription emu or mainland emu. error; other specimens are listed against this number in the old register. Description The femur is somewhat abraded at the extremities, exposing cancellous bone, but

40 3.15 Sandy Cape putative emu vertebra (unregistered)

Plate 24. Sandy Cape putative emu vertebra (top), compared with cervical vertebrae of mainland emu (Melbourne Museum B12822).

Item(s) of a mainland emu specimen on loan from Melbourne Museum (B12822). Based on this 1 x vertebra comparison, the Sandy Cape vertebra is not emu Provenance data – the general morphology of the bone cannot be matched with the distinctive form of emu The vertebra is part of a small collection of vertebrae (Plate 24). bones of various animals from a midden, possibly at Sandy Cape. A partly illegible note dated 5/2/1955 lists the bones with putative identifications. The vertebra is listed as ‘emu ?’. A more recent card label states that the bones were ‘found in a box’ and ‘transferred from Zoology 22/8/2005’. No other details appear to have been recorded. Description The vertebra is very weathered and porous. The spinous process is missing. The colour is greyish white with traces of red mineral or pigment (Plate 24). Dating This specimen is not emu and was not selected for dating. Discussion The identification of this specimen as emu was tested by comparing the bone with that

41 Cosgrove, R, Field, J, Garvey, J, Reid, C, Bertuch, F, Levchenko, 4.0 References Brenner-Coltrain, J, Goede, A, V & Cooper, A 2012, ‘Man and Anonymous 1923, ‘The story of Charles, B, Wroe, S, Pike-Tay, A, megafauna in Tasmania: closing the emu’, The Examiner, 26 July Grun, R, Aubert, M, Lees, W & the gap’, Quaternary Science 1923, p. 4. O’Connell, J 2010, ‘Overdone Reviews, no. 37, pp. 38-47. overkill - the archaeological Gillespie, R, Wood, R, Fallon, Baird, RF 1991, ‘Avian fossils from perspective on Tasmanian S, Stafford, TW & Southon, J the Quaternary of Australia’, in P megafaunal extinctions’, Journal 2015, ‘New 14C dates for Spring Vickers-Rich, JM Monaghan, RF of Archaeological Science, no. 37, Creek and Mowbray Swamp Baird & TH Rich (eds), Vertebrate pp. 2486-2503. Palaeontology of Australasia, megafauna: XAD-2 processing’, Monash University Publications Dickison, D 1926, ‘When did Archaeology in Oceania, vol. 50, Committee, Melbourne, pp. the Tasmanian emu become pp. 43-48. exinct?’, The Emu, no. 25, p. 213. 809-870. Green, RH 1969, The birds of Banks, MR, Colhoun, EA & Dooley, R 2017, ‘From Flinders Island, Records of the van de Geer, G 1976, ‘Late ‘abundance of emues’ to a rare Queen Victoria Museum, no. 34, Quaternary Palorchestes azael bird in the land: the extinction pp. 1-32. of the Tasmanian emu’, Papers & (Mammalia, Diprotodontidae) Gunn, R 1852, Letter, Proceedings Proceedings: Tasmanian Historical from northwestern Tasmania’, of the Royal Society of Tasmania, Research Association, vol. 64, no. Alcheringa, vol. 1, no. 2, pp. 159- no. 3, pp. 168-170. 166. 3, pp. 4-17. Hogg, AG, Hua, Q, Blackwell, Eberhard, R 2014, ‘Dromaius Beruldsen, GR 1980, A Field PG, Niu, M, Buck, CE, Cave: a hitherto unreported Guide to Australian Birds’ Nests Guilderson, TP, Heaton, cave containing bone material and Eggs, Rigby, Adelaide. TJ, Palmer, JG, Reimer, PJ, at Mole Creek’, Speleo Spiel, no. Reimer, RW, Turney, CSM Bronk Ramsey, C 2017, ‘Methods 402, pp. 15-16. for Summarizing Radiocarbon & Zimmerman, SRH 2013, Datasets’, Radiocarbon, vol. 59, Fallon, SJ, Fifield, LK & Chappell, ‘SHCal13 Southern Hemisphere no. 2, pp. 1809-1833. JM 2010, ‘The next chapter Calibration, 0-50,000 years in radiocarbon dating at the cal BP’, Radiocarbon, no. 55, pp. Brown, S 1993, ‘Mannalargenna Australian National University: 1889-1903. Cave. A Pleistocene site in Bass status report on the single stage Hope, JH 1969, Biogeography Strait’, in MA Smith, M Spriggs AMS’, Nuclear Instruments and of the mammals on the islands & B Frankauser (eds), Sahul in Methods in Physics Research B, no. of Bass Strait with an account of Review: Pleistocene Archaeology 268, pp. 898–901. in Australia, New Guinea and variation in the genus Potorous, Island Melanesia, Department of Gill, ED 1953, ‘Distribution PhD thesis, Monash University. of the Tasmanian devil, the Prehistory, Australian National Hope, JH 1973, ‘Mammals of the Tasmanian wolf and the dingo in University, Canberra. Bass Strait Islands’, Proceedings SE Australia in Quaternary time’, of the Royal Society of Victoria, vol. Christidis, L & Boles, WE 2008, Victorian Naturalist, vol. 70, no. 5, 85, no. 2, pp. 163-195. The Systematics and Taxonomy of pp. 86-90. Australia Birds, Horton, DR & Murray, P 1980, Gill, ED & Banks, MR 1956, CSIRO Publishing, Melbourne. ‘The extinct Toolach Wallaby ‘Cainozoic history of Mowbray (Macropus greyi) from a spring Colhoun, EA, Van de Geer, Swamp and other areas of mound in north western G & Mook WG 1982, north-western Tasmania’, Tasmania’, Records of the Queen ‘Stratigraphy, pollen analysis, and Records of the Queen Victoria Victoria Museum, no. 71, pp. 1-12. paleoclimatic interpretation of Museum, no. 6, pp. 1-42. Pulbeena Swamp, northwestern Hume, JP 2017, Extinct Birds, 2nd Gillespie, R, Camens, AB, Tasmania’, Quaternary Research, edition, Bloomsbury Publishing, Worthy, TH, Rawlence, NJ, no. 18, pp. 108-126. London.

42 Hume, JP, Steel, L, Middleton, Scott, HH 1931, ‘The extinct Vickers-Rich, P & Rich T 1993, G, & Medlock, K 2018, In Tasmanian emu’, Papers & Wildlife of Gondwana, Reed, search of the dwarf emu: a Proceedings of the Royal Society of Chatswood NSW. palaeontological survey of Tasmania, pp. 108-110. Williams, DGL 1981, ‘Genyornis King and Flinders islands, Bass Serventy, DL 1967, ‘The Bass eggshell (Dromornithidae; Aves) Strait, Australia, Contribuciones Strait islands’, Australia Natural from the Late Pleistocene of Científiicas del Museo Argentino History, no. 15, pp. 401-408. South Australia’, Alcheringa, vol. de Ciencias Naturales, no. 7, pp. 5, pp. 133-140. 81–98. Spencer, B & Kershaw, JA 1910, ‘A collection of sub-fossil bird Williams, DLG & Vickers-Rich, Legge, WV 1907, ‘The emus of and marsupial remains from P 1991, ‘Fossil Eggs from the Tasmania and King Island’, The King Island, Bass Strait’, Memoirs Tertiary and Quaternary of Emu, vol. 6, pp. 116-119. of the National Museum of Australia’, in P Vickers-Rich, JM Le Souef, D 1904, ‘Extinct Melbourne, no. 3, pp. 5-35. Monaghan, RF Baird & TH Rich Tasmanian emu’, The Emu, vol. 3, (eds), Vertebrate Palaeontology of Sutherland, FL & Kershaw, RC pp. 229-231. Australasia, Monash University 1971, ‘The Cainozoic geology Publications Committee, McIntosh, PD, Price, DM, of Flinders Island, Bass Strait’, Melbourne, pp. 871-891. Eberhard, R & Slee, AJ 2009, Papers & Proceedings of the Royal ‘Late Quaternary erosion events Society of Tasmania, vol. 105, pp. Wylie, J 2013, ‘Scotchtown in lowland and mid-altitude 151-178. Cave, north-west Tasmania: the Tasmania in relation to climate discovery of Tasmania’s first Thomson, VA, Mitchell, change and first human arrival’, cave with an assemblage of KJ, Eberhard, R, Dortch, J, Quaternary Science Reviews, vol. megafauna’, Caves Australia, no. Austin, JJ & Cooper, A, 2018, 28, pp. 850-872. 195, pp. 20-25. ‘Genetic diversity and drivers Patterson, C & Rich, PV 1987, of dwarfism in extinct island ‘The fossil history of the emus, emu populations’, Biology Letters, Dromaius (Aves: Dromaiinae)’, 20170617. Records of the South Australian Turney, CSM, Flannery, TF, Museum, vol. 21, no. 1, pp. 85-117. Roberts, RG, Reid, C, Fifield, LK, Queen Victoria Museum and Higham, TFG, Jacobs, Z, Kemp, Art Gallery 2006, Treasures of N, Colhoun, EA, Kalin, RM & the Queen Victoria Museum and Ogle, N 2008, ‘Late-surviving Art Gallery, Foot & Playstead Pty megafauna in Tasmania, Ltd, Launceston. Australia, implicate human involvement in their extinction’, Ridpath, MC & Moreau, RE Proceedings of the National 1966, ‘The birds of Tasmania: Academy of Sciences, vol. 105, no. ecology and evolution’, Ibis, no. 34, pp. 12150-12153. 108, pp. 348-393. Van de Geer, G, Colhoun, Scott, HH 1915, A note on EA & Mook, WG 1986, ‘“Palorchestes,” as a Tasmanian ‘Stratigraphy, pollen analysis and Pleistocene Genus’, Papers & palaeoclimatic interpretation of Proceedings of the Royal Society of Mowbray and Broadmeadows Tasmania, pp. 100-101. Swamps, northwestern Scott, HH 1923, ‘A note on Tasmania’, Australian Geographer, the King Island emu’, Papers & vol 17, pp. 121- 133. Proceedings of the Royal Society of Tasmania, pp. 103-107.

43 Appendix QVMAG old register Transcriptions of line entries referencing emu material (pages 100-101). Content presented as written:

Newstead House emu leg (QVM.2002.2.1)

Number 1486

Vernacular Name Leg of Tasmanian Emu, with the skin attached – minus the Femur,

Scientific Name Dromeus

Sex [male symbol]

Locality and General History Obtained from the late Ronold Gunn [sic] from St Pauls Plains (bird Donor or Seller including Amount Paid hatched from egg by a Turkey hen) From Newstead house cellar Feb 9.1921. Gunn gave a paper to Royal Remarks Socty of Tas. H Scott also, years afterwards.

Case Number [blank]

Date Acquired 9/2/21

Date Exhibited 1921

Irishtown tibiotarsus (QVM.1990.GFV.144)

Number 1487

Vernacular Name Tibio Tarsus Tasmanian Emu

Scientific Name Dromeus

Sex [blank]

Locality and General History [blank]

Donor or Seller including Amount Paid Given Messers Fenton & Willis

Remarks [blank]

Case Number [blank]

Date Acquired [blank]

Date Exhibited [blank]

44 Mowbray Swamp postcranial bones (QVM.1990.GFV.138-143)

Number 1488

Femur, Tibio Tarsus, 2, Tarso-metatarsi, cervical, + Vernacular name synsacrum of Tas emu.

Scientific name Dromeus

Sex [female symbol]

Locality and general history Purchased from Mr T. Edwards, of Mowbray Swamp.

Donor or seller including amount paid [blank]

Remarks Fairly good order rubbed proximally.

Date acquired Oct 1924

Date exhibited [blank]

Mole Creek tibiotarsus (QVM.1991.GFV.54)

Number 1489

Tibio-Tarsus Vernacular name Tas Emu

Scientific name Dromeus

Sex [female symbol]

Locality and general history From Sassafras Farm, Mole Creek.

Donor or seller including amount paid Given by Mr E.W. Clarke.

Remarks Fairly good order rubbed proximally.

Date acquired 13/3/31

Date exhibited [blank]

45 Pre 1897 femur (QVM.1490)

Number 1490

Femur Vernacular name Tasmanian Emu (?)

Scientific name Dromeus, Sp

Sex [female symbol]

No very exact history – I found in 1897 in the old museum, listed as a Locality and general history “Deers’ bone” to contrast with some

New Zealand Moa remains. It may or may not be Tasmanian - it is Donor or seller including amount paid sightly immature in the sense of being devoid

of all super ossification. I presumed that Alex Morton gave the bone Remarks or acquired it

Date acquired Prior to 1897

Date exhibited 1897, & earlier

Jackson egg (QVM.1965.GFV.0006)

Number 1512

Vernacular name Two small scraps of egg shell of an Emu.

Scientific name [blank]

Sex [blank]

Locality and general history From near Amita [sic] Found by Mr Frank Jackson 600 feet above Donor or seller including amount paid sea level 4 miles inland Remarks Date acquired ap. 10.1917

Date exhibited ap. 11.1917

46 47 55